Replace per-project files with combined data/train.jsonl (509 rows) + dataset card

README.md

@@ -30,9 +30,9 @@ A benchmark for **test-update** in evolving Java codebases. Each row captures a
 
 ## Dataset summary
 
-`teb-update` targets the **positive** subset of TestEvo-Bench: pairs where an existing test method was edited (not added, not deleted) between `rev1` and `rev2`, and the edit is plausibly a real adaptation to a production-code change rather than incidental refactoring. The unit of evaluation is "given the old test, the production-side diff, and dependency context, produce the updated test."
+`teb-update` is a benchmark for updating existing test methods to track changes in the production code they exercise. Every example was mined from a real commit (`rev1` → `rev2`) on a public Java project, and the test edit reflects an adaptation to a behavior- or signature-level change in the focal production method rather than purely cosmetic refactoring.
 
-Each row corresponds to a single (project, `rev1` → `rev2`) pair. Inside the row, `test_changes` is a list of one or more individual test methods that were modified in that revision, each with its own focal-method analysis.
+Each row corresponds to a single (project, `rev1` → `rev2`) pair. Inside the row, `test_changes` is a list of one or more individual test methods that were modified in that commit, each paired with its focal production method, the production-side before/after, candidate dependency methods, and the dependencies that changed in the same commit.
 
 | | |
 |---|---|